Abstract

Stakeholder involvement in environmental management in Australia is increasingly participatory. A community-government partnership has been the central focus of the NSW program of water reforms since the inception of the Water Act 2000. Under the Act, it is the joint responsibility of the New South Wales Government and stakeholder-based Water Management Committees to develop Water Sharing Plans. Water Management Committees typically comprise government, environmental and agricultural representatives. This paper describes a System Dynamics group model-building workshop conducted with the Northern Rivers Water Management Committee. A role of the Committee was to determine the allocation of water between the water users and the environment. In order to achieve this goal, the Committee was provided with hydrological model outputs and other data by the NSW Government, however the decision support framework did not allow the incorporation of existing stakeholder knowledge and understanding. A group model-building workshop was convened to redress this problem. Participants found that the group model-building process was inclusive, enabled the exploration of different understanding, and incorporation of different knowledge-bases. Participants also found the approach allowed a holistic view of the water management system to be developed and assisted participants to understand the inherent complexities. Participants also commented that the group model-building experience assisted in a better understanding of the model development process. The models generated from the workshop and the response from participants suggest that there is potential for wider application and evaluation of System Dynamics group model-building as an extension tool.

Media summary

Keywords

Introduction

Since 1994, States and Territories in Australia have been reviewing their water policies to meet the national Coalition of Australian Governments (COAG) Water Reform Framework. The COAG recognised that, in order to redress the widespread degradation of water resources, a package of measures was required to encompass the economic, environmental and social implications of water usage. The reforms cover all aspects of the water industry, including institutional arrangements for regulation, management and service provision, water allocations and entitlements, water pricing, environmental protection, and emphasise the need for community input into decision making (Cullen 2000).

In New South Wales (NSW), in common with many other states of Australia, there has been a general evolution in all areas of natural resource management, from regulatory administration to outcomes-based legislation and a more participatory approach (Kerin 2000). A community-government partnership has been the central focus of the NSW program of water reforms. Under the NSW Water Act 2000 it is the joint responsibility of the State Government and Water Management Committees to develop Water Sharing Plans. These Water Sharing Plans set in place rules for water use on a catchment scale for instance:

• determining the proportion of flow available for extraction of surface water,
• recommending preferred drought management options, and
• developing local water trading rules.

In order to develop such Water Sharing Plans each Water Management Committee requires a considerable amount of knowledge and understanding of numerous catchment and land use systems. For stakeholder government partnerships, such as Water Management Committees, to succeed in fulfilling their roles, there needs to be good communication between stakeholders, a holistic view (systems thinking), and an adaptive and interactive approach to decision making and decision support (den Exter 2003, Leitch et al. 2001).

The current decision support approach used by the NSW Department of Infrastructure Planning and Natural Resources (formerly the Department of Land and Water Conservation) for preparing Water Sharing Plans in unregulated surface water systems is the ‘Integrated Quantity and Quality Modelling’ (IQQM) framework. The IQQM framework provides stakeholders with a range of information on which to base their decisions. However, because of the technical nature of the modelling, and the number of Water Sharing Plans being prepared, stakeholders are limited to the number of scenarios they can request. Further, it is difficult to introduce new parameters to the IQQM framework, resulting in a lack of input to the model from water management stakeholders. This paper demonstrates the potential of System Dynamics modelling to address such shortcomings.

System Dynamics

System Dynamics has evolved from the combination of two divergent systems approaches–engineering systems, and management decision-making systems (Forrester 1960). System Dynamics is a unique analysis tool in that it forces the description of a system in terms of its structure and function, which in turn generate system behaviour. As its name suggests, System Dynamics modelling is most useful for understanding the behaviour of trends over time. Building System Dynamics models assists in developing hypotheses, determining timeframes (commonly referred to as the ‘reference mode’) and determining the availability of historical data. According to the assumptions of System Dynamics:

• things are interconnected in complex patterns;
• the world is made up of rates, levels and feedback loops;
• information flows are intrinsically different from physical flows;
• non-linearities and delays are important elements in systems; and
• system behaviour arises out of system structure (Meadows 1989, Lane and Oliva 1998).

System Dynamics group model-building

The System Dynamics group model-building process has been described as more of an art than a science (Anderson et al. 1997). Methods used to build group models usually involve conferences (for example Anderson and Richardson 1997) or workshops (for example Maani 2000, Wolfenden 1999). Techniques used within these group model-building forums are varied and rarely well documented (Anderson and Richardson 1997).

The variety of techniques used for group model-building relate to its many purposes and levels of participation. Some cases emphasise the model outcome, for example learning from the insights gained by playing with the model. Others emphasise the process, for example consensus-building and team learning through the development of conceptual models.

Stakeholders may participate in group model-building to different degrees. Stave (2003) identifies a spectrum of participation when involving stakeholders in model-building. At the least participatory level, a completed model may be used to demonstrate the effects of alternative policies to stakeholders. A more participatory approach would allow stakeholders to suggest scenarios to be tested. At the most participatory level stakeholders can help develop the simulation model that represents system structure.

In environmental management, group model-building has been used to involve more than one expert to increase the validity and robustness of a model (Costanza et al. 1990). A recent development has been to use group model-building to involve stakeholders in developing environmental management policy. For example, Wolfenden (1999) and Stave (2002, 2003) have used group model-building to facilitate greater public participation in water planning and decision making. This paper describes three uses for group model-building in water management decision making:

1. Improving a prototype System Dynamics model;

2. Eliciting issues of importance to stakeholders; and

3. Relating issues of importance to stakeholders to the existing decision making framework.

To demonstrate the potential of System Dynamics modelling for improved stakeholder involvement in water management decision making, a case study using the development of the Draft Water Sharing Plan for Coopers Creek, NSW, was undertaken. The Northern Rivers Water Management Committee (the Committee) is responsible for developing Water Sharing Plans for a range of high conservation value catchments in the Northern Rivers region of northern NSW and comprises the stakeholders referred to in this paper. The Committee was formed in 2000 and has twenty-four members representing seven non-government organisations, three local government agencies, six state government agencies and the Northern Rivers Catchment Management Board. The monthly meetings of the Committee from June 2000 to December 2001 form the basis of this research.

It became apparent after attending a number of meetings of the Committee that there were problems with decision support. For example the decision-support information being produced for Committee members (mostly derived from the Integrated Quantitative Qualitative Modelling (IQQM) framework) was not easily related to stakeholder expertise. Further, requested information was difficult to produce due to excessive workloads and the technical nature of the IQQM framework. As a result, much of the Committees time was spent trying to incorporate issues of concern to stakeholders that were not reflected in the IQQM modelling reports. A real-time model at the property scale was required. The researcher (den Exter) recognised that System Dynamics modelling had the potential to provide a flexible and real–time simulation where the existing decision support frameworks could not and set out to test its usefulness to the Committee.

Methods

Research Framework

The evaluation of the use of Systems Dynamics modelling by the Committee was structured around an Action Research Cycle (Figure 1). In this research the Action Research Cycle began by developing key relationships with stakeholders and researching the management system (Step 1). A prototype simulator was developed to assist stakeholders to visualise their existing understanding, and introduce them to the System Dynamics approach (Step 2). The next step in the Action Research Cycle (Step 3) was the group model-building workshop where:

1. The prototype model was presented to Committee members;

2. Options were discussed for the group model-building project and deliberated upon;

3. Issues for each stakeholder were identified;

4. Relationships between the issues were identified and a conceptual model developed; and

5. An evaluation of the process was conducted.

In the final stage of the Action Research Cycle (Step 4) the models developed during earlier steps were reflected upon thus revising understanding of the water management system.

Figure 1: Action Research Cycle

Results

Step 1: Developing key relationships and system understanding

The researcher was a member of the Committee and regularly attended meetings for two years prior to the group model-building workshop. Good working relationships with Committee members were thus established before conducting the exercise. Further, as a stakeholder in the decision making process, the researcher had a good understanding of the water management decision making system and the information available. The development of a prototype model did not take as long as it might have had the researcher not been as involved.

Step 2: The Water Management Simulator Prototype

The models described in this paper were built using iThink software, a third generation computer language. To ensure transparency, comprehension and validity, the prototype Water Management Simulator (WATERSIM) was developed using data already provided to committee members by the NSW Department of Land and Water Conservation (now the NSW Department of Infrastructure Planning and Natural Resources). The researcher transformed this information into a System Dynamics model using her understanding of the processes and linkages built up by attendance at the committee meetings. No new parameters or data, however, were introduced until after the group model-building workshop. The major sectors of WATERSIM are flow, maximum extractable water, flow balance, the ‘cease to pump’ and water sharing rules (Figure 2).

Figure 2: Relationships in the prototype Water Management Simulator (WATERSIM)

Flow balance is influenced by the amount of flow entering the system, the ‘cease to pump’ limit, and the allowable maximum amount of extractable water. These factors are determined by the flow rules. The flow balance has a feedback relationship with the allowable maximum amount of extractable water. Flow in this model is user defined and is actual, or simulated where there are data gaps, daily flow for a wet, dry or average year. The data are derived from the NSW Department of Land and Water Conservation (DLWC)’s Integrated Quantitative Qualitative Modelling (IQQM) data, which were the information base for decision-making by the Committee. The flow balance is a balance between actual flow and water extracted. The ‘cease to pump’ is a trigger level set by the Committee at which flows are considered sufficiently low that pumping must cease. Prior to the development of the water sharing plans the ‘cease to pump’ trigger was visual flow in the creek at a certain gauge.

The main interface that the user sees in WATERSIM is the control panel (Figure 3). It allows the user to choose flow data for a particular year, and define water-sharing rules. Results of the user-defined scenarios are displayed on the graph pad. Variables displayed are flow in (actual flow data (mL/day) selected by the user), environmental flow (flow remaining after extraction in mL/day), the amount of flow able to be extracted (mL/day) and the cease to pump (mL/day) over one year.

Figure 3: WATERSIM control panel. The control panel allows the user to experiment with daily flow data for typical wet, dry or average years, different cease to pump levels ranging from 17mL to 200mL at the end of the system gauge and different water rules (bulk access regimes) for the A, B and C flow classes. The results of each run are displayed on the graph in terms of the daily flow in (determined by the flow data selected), environmental flow (flow left after extraction), extraction (determined by the water rules) and the cease to pump (CTP). The irrigation index is an indicator of when demand is not satisfied (e.g. it flashes red when the peak daily demand is not met). The environmental flow index is an indicator of when environmental flow requirements are not met, for instance in terms of requirements for Eastern Freshwater Cod migration and breeding.

Step 3: Group Model-building Workshop

The System Dynamics group model-building workshop was conducted on the 23^rd August 2002. Between the twelve workshop participants, all member organisations of the Committee were represented at the workshop. Results from the group model-building process are summarised below.

1. Presentation of prototype model

The prototype model (WATERSIM) was presented to the group by the researcher and various decision rule scenarios were examined. Inputs to the WATERSIM model were outputs from NSW DLWC hydrological and irrigation modelling. Thus workshop participants had a good understanding of the assumptions behind the model, and they all had hard copy tables and graphs of the data to refer to at the workshop. The different presentation of this information, however, sparked new questions and discussion on the assumptions behind the modelling.

2. Deciding the model focus

Four options for group model-building were generated from areas of concern previously identified by the stakeholders. These included further development of the WATERSIM prototype, but also included other options so that the facilitator did not dominate the direction of proceedings, but allowed stakeholders to govern the process and hence have greater ownership of and commitment to it. The four options presented were to:

1. continue to examine the water management simulator;

2. develop an issues-based model to examine inter-relationships between stakeholder issues;

3. develop a dairy-farm model to examine the effects of the water-sharing plan on an average farm; or

4. develop a model for the Eastern Freshwater Cod to be used as an environmental indicator.

After some discussion the dairy-farmer model option was chosen. Reasons for choosing the dairy-farm option included participant concern about the effect of the Committee’s recommendations on dairy farmers in the Coopers Creek sub-catchment and the knowledge that the Committee would have little input into the next stage of planning and development of the implementation plan. Developing their own model of dairy farm effects was seen by stakeholders as one avenue in which issues outside the ‘role’ or statutory obligations of the Committee could be examined with relation to the decisions that the Committee had the power to make.

3. Issue identification

Results of the brainstorming process highlighted issues of concern to members of the Committee (Table 1). The group had to decide at this point on the model boundaries. Initially water managed was the main focus but the focus shifted to the modelling the milk production system, the understanding of which was considered critical at a property-scale when properly considering water management options and which had not been considered in the wider process to date (Figure 4). All issues classified as endogenous and exogenous were included in the model developed by the Committee. Factors in the outer ring, although recognised as important effectors of the milk production system, were excluded from the conceptual model as uncontrollable variables.

Table 1: Dairy farmer issues identified by modelling workshop group

Pumping hours Daily access Cut-out time Number of paddocks under irrigation Demand for milk Seasonality

Rostering Resource security Active versus inactive licences Costs – animal health, feed, fertiliser, mgt Water sharing Water users group Water trading Demand for water

Benefits Profit Production system Infrastructure alternatives Management and labour Feed price NORCO benefit Calving relationship

Figure 4: Milk Production System Model boundaries and issues categorisation.

4. Development of Conceptual Model

The initial qualitative model of milk production developed by participants of the workshop contained all the elements considered important by each stakeholder and illustrated the connections between them. Ideas flowed freely from the floor and were recorded on butcher’s paper. The researcher converted the workshop results into the iThink System Dynamics modelling language after the workshop and provided all Committee members with a copy in a summary report. The resulting Milk Production System (MPS) model (Figure 5) remained qualitative at this stage.

Figure 5: The Milk Production System conceptual model developed by stakeholders at the workshop.

The Milk Production System (MPS) model (Figure 5) illustrates how stakeholders perceived links between water management decision-making and the profitability of a dairy enterprise. Thus the consequence of Committee decisions became more realistic for stakeholders. The MPS model also established the role and importance of decisions by the water user group in the determination of daily access to water.

5. Evaluation

Participants were asked to evaluate their experience at the end of the group model-building workshop by anonymously responding to four questions:

5. What was the best/most positive aspect of the exercise?

6. What would you like to have seen done differently?

7. How do you think this type of approach could assist the Northern Rivers Water Management Committee?

8. Any further comments?

Participant responses to these questions are presented in Table 2.

Table 2: Responses to evaluation questions
Question 1: What was the best/most positive aspect of the exercise?	Highlighted problems of water sharing. Highlighted processes, linkages and outcomes. Holistic approach. Use of qualitative factors. Explored different understandings. Brought together different ideas and interpretations. Involved everyone/group participation. Helped understand modelling process and assumptions. Visual aspect of the exercise.
Question 2: What would you like to have seen done differently?	Start simpler before moving to more complex models. Develop model purpose/aim more carefully/specifically. More in depth explanation. Resources: more time, better location, bigger butchers paper, co-facilitator.
Question 3: How do you think this type of approach could assist the Committee?	Useful for communication. Could be useful for decision support but who will do it? Developing a holistic picture where social, economic and environmental issues are all taken into account. Looking at trade-offs. Helping to see interactive and connective nature of issues. Small group work productive. Assists thought processes. Assists in understanding the outcomes of decisions. Helpful in understanding the complexities of a water plan.
Comments	A useful session Willing to do more Relevant to whole committee What kind of commitment is required to get a decision support system developed.

Step 4: Revision of system understanding and prototype model

The original water management simulator (WATERSIM) was revised using the knowledge conceptualised in the qualitative MPS model (Figure 5). In this second version of the water management simulator (Figure 6) a daily access sector was added to the prototype (Figure 2). This allowed for farm-scale analysis of the implications of the different decision options for the ‘cease to pump’ limit at a daily access level and produced more realistic, relevant, and sensitive results than the initial prototype. With further development, the revised WATERSIM model could be used to assist water user groups negotiate and trade water shares, thus reducing negative impacts arising from the water sharing plans.

Figure 6: Relationships between model sectors in the revised WATERSIM model where the allowable maximum extractable water now also influences daily access to water.

Discussion

The prototype model of the water management decision-making system (WATERSIM) provided a good introduction to System Dynamics modelling for Committee members as it illustrated its potential for real-time simulation. As workshop participants had a good understanding of the assumptions behind the data, explaining the parameters and assumptions of the WATERSIM model was much easier than if new parameters had been used. Using parameters that participants were already familiar with also assisted in explaining the language of System Dynamics. As indicated from the evaluation (Table 2), however, more time was needed for participants to understand the structure and behaviour of the model and use it more fully.

The conceptual model developed by stakeholders (Figure 5) demonstrated that useful model products can be developed in a very short time with limited resources using group model-building. The model developed was useful in its own right for stakeholder communication and understanding. Further, the conceptual model developed by stakeholders was easily incorporated into the prototype model (WATERSIM) providing information at the property scale.

At the start of the workshop a number of participants expressed a lack of confidence in their ability to build models despite being familiar with modelling as a decision support tool. These participants quickly overcame their fears to enthusiastically participate in and learn from the exercise. As Stave (2003) found, stakeholders participating in group model-building did not need to know anything about systems modelling to gain system insights.

Participant responses to the evaluation questions were encouraging. Participants found that the process was inclusive and that it enabled the exploration of different understandings. Perhaps most importantly, participants found that they developed a holistic view of the water management system. They also commented that the workshop was useful in facilitating a better understanding of the model development process.

The skills and attitudes of the facilitator are key to successful group model-building interventions (Vennix 1996). In this case study, asking workshop participants to decide on the focus for the model avoided facilitator domination and empowered participants in the process. It was recognised, however, by both the researcher and workshop participants that the group model-building workshop could have been improved with an additional facilitator. An additional facilitator would have enabled the development of the conceptual model using formalised System Dynamics language in real time rather than after the workshop. Certainly, conducting a group model-building workshop with one person facilitating would not have been as successful if there had been any more than the twelve workshop participants.

A number of participants commented on the usefulness of the System Dynamics, group model-building approach, but questioned whether it would be adopted in the longer-term, as it would need commitment and funding. These resource factors could be overcome if the benefits are seen to outweigh the costs of the exercise. Overall the outcomes of this case study suggest that System Dynamics group model-building is a potentially useful extension tool especially where:

1. An understanding of complex trends over time is required;

2. Stakeholder knowledge needs to be integrated with existing decision-making frameworks; and

3. Communication and learning amongst stakeholders is desired.

Further evaluation of group model-building in other settings involving different stakeholders and in comparison with other extension tools is required to determine the transportability of the approach. Establishing relationships with stakeholders and gaining commitment for group model-building interventions, however, can take months and even years (den Exter 2003) making such evaluation difficult. The findings of this research, however, are supported by others observations (for example Maani 2000, Stave 2002, 2003, Vennix 1999, Wolfenden 1999) suggesting group model-building is a valuable addition to the tool-box of the extension officer and facilitator.

Conclusion

To understand and manage environmental systems requires communication and learning between stakeholders and an adaptive approach to decision-making and decision support. This case study demonstrates the potential of System Dynamics group model-building as an extension tool for improving stakeholder involvement in environmental management decision making. It was found that System Dynamics group model-building was useful for developing and improving decision support models and integrating existing information with stakeholder issues and knowledge. There is potential for wider application and evaluation of System Dynamics group model-building as an extension tool.

Acknowledgements

This research formed part of a PhD funded by the CRC for Sustainable Tourism. Professor Glyn Rimmington and Dr David Charles-Edwards provided invaluable co-supervision.

References

Anderson, D.F., Richardson, G.P. and Vennix, J.A.M. (1997). Group Model-building: Adding More Science to the Craft. System Dynamics Review 13(2), 187-201.

Anderson, D.F. and Richardson, G.P. (1997). Scripts for Group Model-building. Systems Dynamics Review 13(2), 107-129.

Costanza, R., Sklar, F.H. and White, M.L. (1990). Modeling Coastal Landscape Dynamics. BioScience 40(2), 91-107.

Cullen, P. (2000). Water Reforms, Science and Ecological Outcomes in Australia. International Water Resources Association 10th World Water Congress, Melbourne (International Water Resources Association).

den Exter, K. (2003 (subm)). Integrating Environmental Science and Management: The Role of System Dynamics Modelling. Phd Thesis. (School of Resource Science and Management. Southern Cross University, Lismore).

Forrester, J.W. (1960). The Impact of Feedback Control Concepts on the Management Sciences.

In ‘Collected Papers of J.W. Forrester’. pp 45-60 (Wright Allen Press Cambridge, MA).

Lane, D.C. and Oliva, R. (1998). The Greater Whole: Towards a Synthesis of System Dynamics and Soft Systems Methodology. European Journal of Operational Research 107, 214-235.

Leitch, A., Bellamy, J., Dale, A. and Pollock, L. 2001. The role of science communication in natural resource planning: A case study in the Central Highlands of Queensland. Exploring beyond the boundaries of extension, Australasia-Pacific Extension Network International Conference 3-5 October 2001, University of Southern Queensland, Toowoomba, Queensland.

Maani, K. (2000). Group Model-building for Consensus Building and Team Learning - A Case Study. Sustainability in the Third Millenium - 18th International Conference of the System Dynamics Society, Bergen, Norway. (The System Dynamics Society).

Meadows, D.H. (1989). System Dynamics Meets the Press. System Dynamics Review 5(1), 68-80.

Kerin, J.C. (2000). Involving the Community in Water Management. International Water Resources Association 10th World Water Congress. (International Water Resources Association Melbourne).

Stave, K.A. (2002). Using System Dynamics to Improve Public Participation in Environmental Decisions. System Dynamics Review 18(2), 139-167.

Stave, K.A. (2003). A System Dynamics Model to Facilitate Public Understanding of Water Management Options in Las Vegas, Nevada. Journal of Environmental Management 67, 303-313.

Vennix, J.A.M. (1996). Group Model-Building: Facilitating Team Learning Using System Dynamics. Chichester, Wiley.

Vennix, J.A.M. (1999). Group Model-building: Tackling Messy Problems. System Dynamics Review 15(4), 379-401.

Wolfenden, J. (1999). A Transdisciplinary Approach to Integrated Resource Management: A Pragmatic Application of Ecological Economics. PhD Thesis. (Centre for Water Policy Research, University of New England Armidale).